ziyu3141/rf_newconcat_test1

该数据集包含了图像文件的相关信息，如文件名、审美评分、瑕疵评分、对齐错误评分、总分等。此外，还包括了瑕疵热图和对齐错误热图，以及对齐错误的标记信息。数据集还包含了图像是否均匀、首选图像、非首选图像、修订图像、修订前后的ID标识，以及是否被选为首选图像的标记。整个数据集分为训练集，其中包含了大量的图像数据。

The dataset contains information related to image files, such as filenames, aesthetic scores, artifact scores, misalignment scores, overall scores, artifact heatmaps, misalignment heatmaps, misalignment token labels, information on whether the image is uneven, preferred images, unpreferred images, revised images, IDs of unrevised and revised images, and a boolean indicating whether an image is preferred. The dataset is split into a training set, which contains a large number of image data.